p. 199; vol. xxxv. p. 462; vol. xxxvi. p. 492; vol. xl. p. 317; vol. xlii. p. 80; vol. xlix. p. 259. Portable riveting machines: vol. xxxix. pp. 471, 474; vol. xliv. p. 289 (vol. xliv. p. 299, pneumatic); vol. xlix. p. 256. Hydraulic riveter with plate-closing arrangements, vol. xliii. p. 531. Hydraulic riveter for using rods instead of rivets, vol. xlix. p. 533. Portable riveter for the circumferential seams of boilers, vol. xliii. pp. 490, 491, vol. lxii. p. 422.

The riveting of the last circumferential seams can be performed by the last-mentioned tool when both end plates, having flanges turned inward, have been fitted in place. The front tube plate is made in three pieces, of which the centre one is not fitted till later; the boiler is laid on its back end, and the furnaces, which are not yet riveted to the fronts, are dropped to the bottom, and can, as occasion requires, be shifted to make room for the riveter. This consists of two long arms, of which one passes through the opening of the tube plate to the circumference; the other reaches the circumference from the outside, and contains the hydraulic cylinder, &c.

Another machine (fig. 243), designed for the same object, is worked by steam. To the centre (C) of the end plate is bolted a long arm with a powerful spring and a heavy holding-up weight (W), which presses against the rivet in the circumference of the end plate flange; a strong lever (not shown) is fitted, with which this weight can be pulled back and shifted to the next hole, where the rivet has previously been inserted from inside. The lever is now released, whereby the rivet is pressed home; a small steam hammer (H) is then made to strike the other end of the rivet, shaping it as required.

In works where either of these tools is used-where, therefore, both end plates can be riveted by machinery-the shell seams might have been riveted with the help of comparatively light machines, similar to those used in the construction of the Forth Bridge.

Another plan, which enables all the circumferential seams to be riveted by machinery, is shown in fig. 244, which represents a section through the lower front end

square inch of section of rivet hole seems to be ample, while anything above 150 tons seems to be dangerous. Allowing for the size of the rivet head, and assuming the red-hot rivet to be as fluid under pressure as water, this pressure would give rise to a tension of 50 tons

per square inch in the metal surrounding the hole. Even allowing 50% for friction and other causes, the stress is still an excessive one. The distance of the metal from rivet hole to edge of plate must be sufficient to prevent bulging. On p. 220 it has been shown that, as regards strength of joint, it is necessary that the above dimension d p-n. should be at least equal to where p is the pitch of rivets in N.

the outer row, N the total number of rivets in one pitch, and n the number of rivets in the innermost row.

m

FIG. 245

This leads to the generally adopted practice of making this margin equal to the rivet diameter. The shearing stress which would be set up in this part of the seam, while riveting with a pressure of 150 tons per square inch of rivet hole, would be about 20 tons, which is also a dangerous amount, and the edges of nearly all machine-riveted butt straps will be found bulged.

It must not be forgotten that, as at present constructed, hydraulic riveters exert a greater pressure than their nominal one, for on opening the valve to the cylinder the accumlator weight falls, being arrested only when the rivet is struck. Of course the acquired energy of the drop makes itself suddenly felt as a very serious, but at present not measured, blow, exceeding by many tons the nominal pressure.

It is also important to see that the various plates are screwed close together; otherwise, and particularly if the heads are small or countersunk and the rivets hot, plastic metal will force its way between the plates, as shown in fig. 245. Some riveting machines are so arranged that they can exert a pressure on the surrounding plates before and during the time that the rivet is being pressed. But whether this machine be used or not, it is always good to have the plates well bolted together: if possible, there should be one bolt in every alternate hole.

Bolting Rivet Seams.-The bolts may remain in place till all the alternate holes are riveted up, or they can be gradually removed while the holes are being successively filled. The latter plan seems to be both slower and more unsatisfactory than the first one, for it tends to stretch and shift the plates, and is adopted for this purpose in shipbuilding when the butts do not meet. It is claimed that this plan has the advantage of warming the plate round the succeeding hole by the previous rivet, thereby preventing cracks; but such a danger ought not to exist with good material.

It will be found that the work can be done both better and quicker if the rivet holes are not filled up in succession, for, in order to do this, the bolts near the riveting machine have to be removed, and that, of course, is inconvenient. If the rivets are put slightly into the holes, and at once riveted up, the boiler shell has to be shifted backwards and forwards over a distance of about 2 ft. before the next rivet can be inserted. If, on the other hand, three or four rivets are placed

simultaneously into adjoining holes, the speed of riveting can certainly be increased, but that means that the second rivet is being pressed, while the first one is still hot and pliable, and it will almost certainly be stretched a little by the spring of the plate, so that some of the . beneficial influences of the powerful hydraulic pressure are being wasted. On the other hand, if the pressure is kept up till the first rivet is cold there is no advantage in having three others waiting all this time in

d

a

FIG. 246

adjoining holes. This practice is illustrated in fig. 246. The rivet d is being closed, and e, f, g, h are waiting, while a, b, c are finished. In fig. 247 the rivet b is being subjected to the hydraulic pressure; a has just been pressed, but cannot well be affected by the new operation on account of a strong bolt filling the intermediate hole. The rivet c has just been placed in its hole, and can readily be acted upon when the pressure on b has lasted long enough. When every alternate hole

has been filled, all the bolts are removed together, and their holes are then riveted. After the bolts have been inserted, and before riveting is commenced, hydraulic pressure is applied all round the seam, and if this should slacken some of the bolts they are tightened up.

Irregularly-shaped Rivet Head. Anybody watching the process of riveting must be struck with the primitive means used for guiding the rivet holes to the dies. An overhead crane, which might be more usefully employed, is carrying the boiler, which is constantly swinging about. Crowbars are stuck into some of the rivet holes, and the men tug at these till the rivet, which has been inserted, is in the right position. As the weight of the boiler shell is often nearer 20 tons than 10 tons, it is natural that there must be much pulling and shoving till the right point is approximately reached. But even then the angular position of the shell is not correct, and the centre line of the riveter does not coincide with the axis of the rivet hole. Naturally, when the pressure is now applied, a few small but violent oscillations

take place, and the chances are very much against the rivet head having been formed centrally round its shank. This is illustrated in fig. 248. Of course, externally there is no indication of this state of affairs, except perhaps that the dies leave semicircular marks

round the heads, as shown in fig. 249. It ought not to be a difficult matter to devise appliances

which would turn the boiler shell to its right position with more certainty than is at present possible.

FIG. 249

Time required for Riveting.-When the pressure has been applied it should not be taken off again in less than about one minute (and the longer it remains the better); otherwise the rivet will not have cooled sufficiently, and the remaining spring in the plates will stretch it and reduce its diameter. The men do not generally care to keep the pressure on for so long, because the dies, &c., get hot and soft and soon wear out. It takes about 15 minutes to close up 10 rivets.

The Heating of the Rivets is done in a small reverberatory furnace. Sometimes gas or oil is used as fuel. They should not be raised to too high a temperature, or else they grow too plastic under hydraulic pressure and spread out between the plates (fig. 245, p. 252). Nor should they be heated too quickly, or else only their outsides are softened. On the other hand, it is not good to spend too much time over the heating, for a long exposure, particularly in an oxidising flame, reduces the strength of the material of the outer surface, and it is just this which is subjected to the severest stresses when under working condition. Fifteen to twenty minutes for heating is the general practice. Unlike iron rivets, steel ones cannot be burnt without showing it when in place.

The influence of heating rivets to a red or a white heat is discussed by M. Considère, An. Pont. Ch.,' 1885, 6th ser. vol. ix. p. 574, and 1886, 6th ser. vol. xi. p. 5.

Rivets are made about in. less in diameter than the holes they are intended to fill. The length, L, of their shanks has, therefore, to be made greater than T, the combined thickness of the plates. L=1.5 D +T (1+.D), where D is the diameter of the rivet hole. The weights of rivets may be calculated by the formula

Q. cwt. 2. n.

D2

1000 (L+1.6 D),

where n is the number of rivets.

Internal Parts of Boilers.-Nearly all internal plates, as well as the end plates, have to be flanged. The only operations which precede this one are the bending and riveting or welding of the furnaces, the thinning or drawing out of some corners, and the drilling or punching of a few holes for temporarily securing the plates. Welding will be touched upon later, and the other preliminary operations require no remark, except, perhaps, that the drawing out of the corners may be done either by a steam hammer or by hand. The latter plan takes longer, but seems to produce a better job. (See p. 249.)

Before discussing the various flanging operations it is well to have a clear idea as to the risks attending them.

Dangers attending Flanging Operations. Whenever a piece of iron or steel is being heated there is, of course, a danger of burning it. This should never happen while flanging, for during this operation such a heat ought not to be approached. A more real trouble is the wasting away which takes place, particularly with iron, which loses a considerable part of its thickness each time that it is re-heated. There is the further danger of reducing the thickness of the plate by drawing it out, and also the difficulty of producing the correct shape of flanges, and of preserving it, while adjoining parts are being heated.

More serious than any of these troubles is the risk of cracking the plates. There seem to be three ways of doing this:-1st, by using redshort iron or steel; 2nd, by not annealing the flanged plates, which then retain strains that may ultimately lead to ruptures; 3rd, the working of iron or steel at a blue heat. The plates which are thus injured will either break at once or, being now in a brittle condition, will crack later on.

On all these subjects interesting information will be found in the chapter on Strength of Materials."

The Danger of not Annealing flanged plates has been demonstrated over and over again in boiler yards by plates cracking. Few of the manufacturers care to give details of their experience, especially as the steel makers readily substitute new plates for the spoilt ones, and do not care to have such cases noised abroad. Those which have been published will be found in the chapter on 'Strength of Materials,' but numerous others are continually occurring. Flanged tube plates (figs. 250, 253, 256), flanged end plates (fig. 255), furnace fronts (fig. 251), and with ships the huckster plates, boss plates (fig. 254), and garboard strakes, sometimes give trouble by cracking.

is

Stresses Due to Flanging.-Here it will not be out of place to draw attention to the fact, which is thoroughly supported by experiments, that iron and steel are affected in their quality by severe stresses, and particularly that, by subjecting test pieces to a slowly increasing but severe pull, their ultimate strength can be raised, while their ductility very much reduced. The most important point is that the limit of elasticity rises under this stress, until it actually exceeds the original strength of the material. In other words, taking a test bar whose limit of elasticity is 10 tons, and whose ultimate strength is 30 tons, with 20% elongation, it is possible, by slowly increasing the load, to raise both the limit of elasticity and the ultimate strength to about 35 tons, while the elongation which takes place during the final loading